Early identification of undiagnosed HIV infection is a critical public health priority. In the United States, despite several substantial HIV-related publc health initiatives, approximately 250,000 HIV-infected individuals remain undiagnosed and 50,000 new infections occur annually. Although HIV testing is an important intervention, controversy still exists as to how it should be implemented. In 2006, the Centers for Disease Control and Prevention (CDC) recommended nontargeted opt-out HIV screening in clinical settings where the undiagnosed prevalence was 0.1%. Emergency departments (EDs) have been a major focus of these recommendations, prompted by the fact that over 120 million ED visits occur annually in the United States, they serve large proportions of underserved patients, and are the most common site of missed diagnostic opportunities for HIV infection. In contrast, in 2007 the United States Preventive Services Task Force recommended targeted HIV screening (i.e., testing high-risk subpopulations) as the principal approach to HIV testing because insufficient evidence existed to support the CDC recommendations. Led by Jason Haukoos, MD, MSc, our research team has pioneered investigations in this area since 2004, recently publishing the largest clinical trial to date, concluding that nontargeted opt-out rapid HV screening in the ED was associated with a small increase in number of newly-identified HIV-infected patients when compared to diagnostic testing (i.e., testing based on clinical signs or symptoms) by clinicians. Our team also recently developed the Denver HIV Risk Score (DHRS), the first multivariable tool to estimate risk of HIV infection. The DHRS combines 3 demographic and 5 behavioral characteristics, and classifies patients into distinct strata with increasing HIV prevalence. To build on this work, we propose the following specific aims: (1) to evaluate and compare the effectiveness of 3 modern rapid HIV screening strategies when fully-integrated into routine ED care; (2) to measure and compare program costs of each HIV screening strategy; and (3) to measure and compare ED operational processes of each strategy. In doing so, we will perform a multi-center prospective clinical trial using a quasi-experimental equivalent time-samples design to test the following hypotheses: (1) targeted rapid HIV screening using the DHRS to identify high-risk patients is significantly associated with new HIV diagnoses when compared to traditional targeted rapid HIV screening and nontargeted rapid HIV screening; (2) targeted rapid HIV screening is more cost effective per newly-identified patient than nontargeted rapid HIV screening; and (3) targeted rapid HIV screening does not significantly impact ED processes of care when compared to nontargeted rapid HIV screening. We will leverage an outstanding interdisciplinary research team to conduct the largest and most comprehensive evaluation of HIV screening in EDs. The results will substantially improve our understanding of how to provide effective and efficient rapid HIV screening in EDs and to predominantly underserved populations, and will inform similar practices in other high-risk care settings.